【摘要】 由于遥感技术应用领域的不断扩大,用户希望利用遥感技术来实现功能性用地的自动提取.然而,在现行分类标准中,存在许多按应用功能分类的地物,其在光谱和纹理等遥感特征上与普通的城市建筑物没有本质区别,仅仅通过遥感数据本身,无法对其进行提取.因此,文中以教学用地为例,提供一种用地理本体和相对高程对其进行识别分类的方法:首先,在遥感影像中提取操场和建筑物对象,通过阴影计算建筑物的相对高程并确定其建筑面积,再通过遥感领域知识从操场面积推算出学校的建筑面积,并通过缓冲区分析对教学用地进行分类.更多还原

【Abstract】 As the application field of remote sensing technology continues to expand,the user wants to achieve the automatic extraction of the functional land with remote sensing technology.However,in the current classification standard,many surface features are classified by their function,they have no substantial difference with general urban residential land and other buildings in cities in remote sensing characteristics,such as spectral and texture.Just by remote sensing data itself,we can′t extract them.Therefore,taking teaching land as an example,this article provides a method to identify and classify they by using the geographic ontology and relative elevation.Firstly,the playground and building objects in remote sensing image is extracted to calculate the relative elevation of the building according to its shadow and determine its construction area.Then the school′s construction area is calculated through the playground area with remote sensing domain knowledge.And the classification of teaching land via buffer analysis is realized.It explores a new way for automatic recognition and classification of such targets of functional land.更多还原